Well, maybe you misunderstood my intentions...
What I wanted was to SHARE my opinion on this variation. As I already said I have worked this variation for a long time and so I am aware of its nuances quite well
And well, with my new posts, you had to seek improvements almost everywhere !
Also, I am very critical about computers, but sometimes they are helpful (sometimes not, especially on some endgames...)
Anyway, please believe me that the refutation of your idea 13...Rd8 14.Rb1 b6 15.Qa4 Bd7 16.Bb5 a6 17.Bxc6 Qxe2 18.Rxb6 I could find by myself (I am not so bad

)
Of course, your variation is playable, but this is the case of almost every variation

What I meant was that if white plays well, he endures a durable edge.
I will answer some of your ideas here, but for more details you can check the pgn :
In Sakaev's and Vachier-Lagrave's endgame, I think that you were a bit quick : 17...h5 18.f3 h4 19.Nf1! (a move that the computer doesn't like, by the way

) when white's king comes to e2 and the knight to e3, and white enjoys a durable edge with the weakness remaining on h4 or h3. I don't know the analysis of Sakaev after 17...b6, but I simply don't see how white will be better.
In what I think is the main variation, after your new idea 15...Na5 I play 16.Bb5 (because white's knight is quite fine on e2, at least compared to the queen on h5

) and then black's pieces are very badly coordinated.
Your idea in the Atalik ending is not good : after 17...Rb8 18.Rc1 0-0 I play 19.f3 (no more back rank problems !), and I think that black is on the edge of losing !
So you have to play this 4-3 rook ending which you "should try really hard to lose" but I just don't want to count the number of GM's that lost it, including Kasparov (he was in a bad day, of course, but still it means that it is not as easy as you want to believe)
And at least it shouldn't be your goal to get this position on the board.
FInally I wanted to say that everywhere I criticized your ideas it was because I had a different evaluation of the position

and not just because I wanted to criticize...
I know that it is impossible to always give the best moves, but it is important to try to do so in critical situations, and that is the problem of Grunfeld : there are many ones !
Ametanoitos wrote on 06/11/08 at 15:35:42:
Quote:I just wanted to connect our powers to conclude if we can trust, or not this variation.
Well I wanted to help for that, and my general opinion is that in practical play this is fully playable (because of the surprise) but I believe that it is not a reliable variation from a theoretical point of view.
NB : I hope that your "a 2200+ player" and "many hours of work together with strong players (even GMs)" were not meant as authority arguments, because I would trust nobody, including myself, without a deep analysis (and correct of course...)
What would have been the use of telling you that I'm a 2400+ player, and that I also checked this variation with GM's ? I think it would just have sterilized the discussion, although I wish it would not have !
Best regards,
darkness